User Guide v1.0

Documentation

The complete guide to mastering TalkTyper - your intelligent voice-to-text companion.

Introduction

TalkTyper is a powerful macOS menu bar application that converts your speech to text with intelligent processing options. Whether you need simple voice-to-text transcription or AI-enhanced formatting, TalkTyper provides flexible modes to suit your workflow.

Key Features

  • Local & Cloud Processing: Use offline Whisper transcription or cloud providers for maximum flexibility.
  • 7 Intelligent Modes: From basic transcription to AI-powered email and code assistance.
  • Global Hotkeys: Record from anywhere with customizable keyboard shortcuts.
  • Privacy-First: Local processing by default, with optional context capture and full transparency.
  • History & Export: Review, search, filter, and export past transcriptions.
💡

Quick Start

Press ⌃+âŒĨ+Space anywhere on your Mac to start recording. Your transcription will be automatically typed into the active application.

Getting Started

Installation

  1. 1

    Download TalkTyper

    Get TalkTyper from the Mac App Store.

  2. 2

    Open the App

    TalkTyper will appear as an icon in your macOS menu bar.

  3. 3

    Complete Onboarding

    Follow the setup wizard to configure permissions and preferences.

    • Grant Accessibility permission (required for global hotkeys)
    • Grant Microphone permission (required for recording)
    • Configure your preferred hotkey
Download from App Store

First Launch

On first launch, TalkTyper automatically configures sensible defaults:

  • Appears as an icon in your macOS menu bar
  • Uses the default "Voice to Text" mode
  • Uses local Whisper processing (no API keys required)
  • Sets default hotkey to ⌃+âŒĨ+Space
ℹī¸

Permissions Required

TalkTyper requires Accessibility and Microphone permissions to function. You'll be prompted to grant these during first launch.

Main Interface

TalkTyper Main Menu Interface
Main dropdown menu

Menu Bar Icon

The TalkTyper icon in your menu bar provides visual status feedback:

  • Default State: Ready to record
  • Recording: Animated indicator with pulsing red dot
  • Processing: Processing animation while transcribing

Main Menu Elements

Header Section

Displays "TalkTyper" with Pro badge and status indicator.

Mode Selector

Grid of available modes with Pro and AI badges.

Recording Section

Large Record/Stop button with mode display and hotkey badge.

Quick Actions

History and Settings buttons for quick access.

Recording & Transcription

TalkTyper Recording Interface
Recording interface

Starting a Recording

You can start recording in three ways:

  • Menu Bar: Click the microphone button
  • Global Hotkey: Press ⌃+âŒĨ+Space
  • Recording Window: Use dedicated controls

During Recording

  • The menu bar icon shows an animated recording indicator
  • Live duration is displayed in real-time
  • Waveform visualizes your audio levels
  • Real-time transcription appears (if enabled)

Transcription Pipeline

  1. 1

    Audio Processing

    Audio is captured and prepared for transcription. Noise reduction is applied if enabled.

  2. 2

    Speech-to-Text

    Audio is processed by your configured STT provider (Local Whisper or Cloud).

  3. 3

    AI Enhancement

    If using an AI-powered mode, the text is processed by your LLM provider.

  4. 4

    Output Delivery

    Final text is copied to clipboard and/or auto-pasted into the active application.

Intelligent Modes

TalkTyper includes 7 built-in modes designed for specific use cases. Each mode transforms your speech into optimized output.

Voice to Text

Free

Fast, raw transcription with minimal formatting. No AI processing required. Best for quick notes, dictation, and when you need exact speech capture.

Super

Pro

Context-aware AI assistant that adapts to your current application, selected text, and clipboard contents. Intelligently formats output based on where you're working.

Coding

Pro

Transforms voice into structured prompts optimized for AI coding assistants like Cursor, Copilot, and Claude. Extracts requirements, technical details, and intent.

Message

Pro

Clean, casual messages for Slack, Discord, iMessage, and chat apps. Removes filler words, fixes grammar, and maintains conversational tone.

Email

Pro

Professional emails with proper structure. Automatically adds appropriate greetings, organizes content into paragraphs, and includes sign-offs.

Note

Pro

Structured notes with Markdown formatting. Creates headings, bullet points, and organized sections. Ideal for brainstorming, meeting notes, and documentation.

Meeting

Pro

Produces comprehensive meeting summaries including key discussion points, decisions made, and action items with assignees.

Custom Mode Settings
Custom mode settings

Custom Modes (Pro)

Pro subscribers can create unlimited custom modes with personalized prompts, context settings, and provider choices. Duplicate any built-in mode as a template.

Settings

General Settings
General settings

General Settings

Appearance

  • Show in Dock/Menu Bar
  • Theme (System/Light/Dark)
  • Launch at Login

Behavior

  • Default Mode selection
  • Auto-paste behavior
  • Notification preferences
Audio Settings
Audio settings

Audio Settings

  • Input Device: Select your preferred microphone
  • Input Level: Adjust gain from 0x to 2x
  • Microphone Test: Verify your setup is working
  • Sound Feedback: Enable/disable recording sounds
Keyboard Shortcuts Settings
Shortcuts settings

Keyboard Shortcuts

ActionDefault
Start/Stop Recording⌃âŒĨSpace
Cancel RecordingESC
Open Settings⌘,
Open History⌘H
Quit TalkTyper⌘Q
💡

Mode-Specific Shortcuts

Pro users can assign unique hotkeys to individual modes for instant access.
Vocabulary Settings
Vocabulary settings

Vocabulary Settings

Custom Words

Add technical terms, names, or acronyms that TalkTyper should recognize correctly.

Text Replacements

Define auto-corrections like "gonna" → "going to" for cleaner output.

Privacy Settings
Privacy settings

Privacy Settings

  • Context Capture: Enable/disable clipboard and selection context for AI modes
  • History Retention: Control how long transcriptions are stored
  • Data Export: Export or delete all your data
Advanced Settings
Advanced settings

Advanced Settings

  • Mini Window: Configure behavior of the floating recording window
  • Real-time Transcription: Toggle live transcription preview
  • Local Model: Configure CPU threads and GPU acceleration
  • Storage Limits: Set maximum history entries

Provider Configuration

Provider Settings
Provider settings

Configure your Speech-to-Text and LLM providers to customize how TalkTyper processes your audio and generates enhanced output.

Speech-to-Text Providers

Local Whisper (Default)

100% offline processing on your device. No API key required.

  • Model sizes: Tiny, Base, Small, Medium, Large, LargeV3
  • Optimized for Apple Silicon

Cloud Providers (Pro)

Fast, accurate cloud transcription. Requires your own API keys.

  • OpenAI Whisper API
  • Deepgram
  • AssemblyAI

LLM Providers

Cloud Models

  • OpenAI (GPT-4o, GPT-4, GPT-3.5)
  • Anthropic (Claude 3.5 Sonnet, Claude 3 Opus)
  • Groq (Llama 3, Mixtral)

Local Models

  • Ollama - Run open source models locally
  • LM Studio - User-friendly local inference

Automation

Intelligent Mode Switching Settings
Automation settings

Intelligent Mode SwitchingPRO

TalkTyper can automatically switch to the most appropriate mode based on which application you're currently using. This eliminates the need to manually change modes when switching between tasks.

💡

How It Works

When enabled, TalkTyper detects your active application and automatically selects the optimal mode for that context. For example, recording while in Messages will use Message mode, while VS Code triggers Coding mode.

Default App-to-Mode Mappings

TalkTyper comes with pre-configured mappings for popular applications:

Message Mode

  • Messages
  • Discord
  • Slack
  • WhatsApp
  • Telegram

Meeting Mode

  • Microsoft Teams
  • Zoom
  • Google Meet
  • Webex

Coding Mode

  • Visual Studio Code
  • Cursor
  • Xcode
  • Terminal

Email Mode

  • Mail
  • Outlook
  • Gmail (in browser)
  • Spark

Customizing Mappings

You can customize which mode is triggered for each application:

  1. 1

    Open Automation Settings

    Navigate to Settings → Automation tab.

  2. 2

    Enable Intelligent Mode Switching

    Toggle the feature on to activate automatic mode detection.

  3. 3

    Modify App Mappings

    Click on any application in the list to change its associated mode, or add new applications.

ℹī¸

Override Behavior

If you manually select a mode before recording, that selection takes precedence over the automatic mapping for that session. The next recording will resume automatic switching.

History

History Interface
History interface

Features

  • Chronological List: View transcriptions with timestamps and previews.
  • Search & Filter: Find by content, Favorites, Date, or Mode.
  • Quick Actions: Copy, Favorite, Delete, or Reprocess.
  • Reprocess: Run old recordings through different modes.
  • Context Snapshot: See which app was active during recording.
ℹī¸

Export Options

Export your transcription history in plain text, Markdown, or JSON format for backup or integration with other tools.

Subscription & Pro

Subscription Management
Manage subscription

Managing Your Subscription

Access your subscription details from Settings → General, where you can view your current plan, manage billing, and upgrade or modify your subscription at any time.

  • View current subscription status
  • Upgrade from Free to Pro
  • Manage payment methods
  • View billing history

Free Tier

  • Voice to Text mode only
  • Local Whisper processing
  • 10 transcriptions per day
  • 1-minute recording limit

Pro Tier

  • All 7 Intelligent Modes
  • Unlimited transcriptions
  • No recording time limits
  • Cloud & Local provider support
  • Custom Mode creation
  • Context Capture features

Feature Comparison

FeatureFreePro
Voice to Text Mode✓✓
AI-Powered Modes—✓
Local Whisper✓✓
Cloud STT Providers—✓
LLM Providers—✓
Custom Modes—✓
Context Capture—✓
Daily Limit10Unlimited
Recording Time1 minUnlimited

Privacy & Security

🛡ī¸

Privacy-First Design

TalkTyper is designed with privacy as a core principle. By default, all processing happens locally on your device.

Local Processing

With Local Whisper, your audio is processed entirely on your Mac. No audio data is ever sent to external servers.

Secure Storage

All transcription history is stored locally. API keys are stored securely in the macOS Keychain.

Cloud Processing

If you opt to use cloud providers, data is sent only to that specific provider according to their privacy policy.

Context Control

Context capture (clipboard, selection) is optional and clearly labeled. You control what data AI modes can access.

⭐

Your Data, Your Control

You can export or delete all your data at any time from Settings → Privacy. TalkTyper does not collect telemetry or share your data with third parties.

Troubleshooting

Microphone Not Working

Recording doesn't capture audio or the waveform appears flat.

  • Check microphone permission in System Settings → Privacy & Security → Microphone
  • Verify the correct input device is selected in TalkTyper Settings → Audio
  • Test your microphone with the built-in test function in Audio settings
  • Try unplugging and reconnecting external microphones

Global Hotkey Not Responding

Pressing the configured hotkey doesn't start recording.

  • Verify Accessibility permission is granted in System Settings → Privacy & Security → Accessibility
  • Check for conflicts with other applications using the same hotkey
  • Try setting a different hotkey combination in Settings → Shortcuts
  • Restart TalkTyper after granting permissions

Text Not Auto-Inserting

Transcription works but text doesn't appear in the target application.

  • Ensure Accessibility permissions are enabled for TalkTyper
  • Grant Automation permission in System Settings → Privacy & Security → Automation
  • Some applications may block automated typing - try copying to clipboard instead
  • Check if "Auto-paste" is enabled in Settings → General

Slow Transcription

Processing takes longer than expected.

  • Local Whisper processing depends on your Mac's hardware - Apple Silicon is recommended
  • Try a smaller Whisper model (e.g., Tiny or Base) in Provider settings
  • For faster results, consider using a cloud STT provider
  • Close resource-intensive applications while transcribing

API Key Errors

Cloud provider returns authentication or rate limit errors.

  • Verify your API key is correct and has not expired
  • Check that your API key has the required permissions for the service
  • Ensure you haven't exceeded your API rate limits or quota
  • Try regenerating your API key from the provider's dashboard

Frequently Asked Questions

By default, TalkTyper uses Local Whisper processing, which means your audio never leaves your device. If you choose to use cloud providers (OpenAI, Deepgram, etc.), audio is sent only to that specific provider according to their privacy policy.
For local processing with Whisper, no internet is required. Cloud STT and LLM providers require an internet connection. Voice to Text mode with local Whisper works completely offline.
Free tier includes Voice to Text mode with local Whisper processing, limited to 10 transcriptions per day with 1-minute recordings. Pro unlocks all 7 modes, unlimited transcriptions, no time limits, cloud providers, and custom mode creation.
Yes! TalkTyper supports bringing your own API keys for OpenAI, Anthropic, Groq, Deepgram, and AssemblyAI. Keys are stored securely in the macOS Keychain.
Yes, TalkTyper is optimized for Apple Silicon (M1/M2/M3/M4) and takes advantage of the Neural Engine for faster local AI processing. Intel Macs with AVX2 support are also supported.
Pro subscribers can create custom modes in Settings → Modes. You can duplicate an existing mode as a template, then customize the prompt, context settings, and provider choices.
TalkTyper supports 99+ languages through Whisper, including English, Spanish, French, German, Chinese, Japanese, Korean, Arabic, Hindi, and many more. Language detection is automatic.
Yes, you can export your history in various formats including plain text, Markdown, and JSON. Access this through the History panel's export function.

Appendix

Supported Languages

TalkTyper supports transcription in 99+ languages through Whisper, including:

English
Spanish
French
German
Italian
Portuguese
Chinese
Japanese
Korean
Russian
Arabic
Hindi

System Requirements

  • OS: macOS 14.0 (Sonoma) or later
  • Processor: Apple Silicon (M1/M2/M3/M4) or Intel with AVX2
  • RAM: 8GB minimum (16GB recommended for local AI)
  • Disk Space: 500MB - 3GB (depends on model size)

Keyboard Modifier Reference

Command⌘
OptionâŒĨ
Control⌃
Shift⇧